Extracellular proteoglycans: A multiscale computational study

The extracellular matrix (ECM) plays an important role in cell migration and proliferation, and hence in the progression of malignant tumors. Healthy brain is one of the softest tissues, but in cancer, the stiffness of the brain ECM can increase by one or two orders of magnitude. Within the framework of studying mechanotransduction in brain tumors, we aim to construct a multiscale computational model of the brain ECM. In contrast to most tissues whose mechanical properties are governed primarily by collagen, the brain ECM is mainly composed of proteoglycans and hyaluronic acid. During cancer, expression of different proteoglycans is altered during remodeling of the brain ECM.

Proteoglycans are composed of a core protein to which a number of glycosaminoglycan chains are attached. One of the best characterized proteoglycans is aggrecan, the key ingredient of cartilage. The biophysical properties of proteoglycans such as brevican, neurocan and versican, which are predominant in the brain ECM, however, are much less known. In this study, we use computational modeling to predict the biophysical properties of these proteoglycans. Due to the large scale of these macromolecules (with molecular weights of a few MDa) coarse-grained models are needed to model proteoglycans. In this work, we combine a one-bead-per-aminoacid (1BPA) model, developed by Onck and coworkers, to represent the core protein with our recently developed one-bead-per-saccharide (1BPS) model for modeling glycosaminoglycans. The combination of these two coarse-grained models allows us to predict the properties of proteoglycans as a function of the length of the side chains, grafting density, salt concentration and length of the backbone. Validation of the model is performed with respect to the biophysical properties of aggrecan.